Centralized machine learning algorithms in vehicular networks face privacy and resource constraints. Federated Learning (FL) addresses these by enabling collaborative model training without sharing raw data. To incent...
详细信息
The ability of Convolutional Neural Networks (CNNs) to accurately discriminate between normal and tumorous brain tissues has been promising. The review focuses on the different CNN models, pre-processing methods, data...
The ability of Convolutional Neural Networks (CNNs) to accurately discriminate between normal and tumorous brain tissues has been promising. The review focuses on the different CNN models, pre-processing methods, data augmentation, and Transfer Learning (TL) strategies used in this research. This Systematic Literature Review (SLR) collected the data from Google Scholar. The results of this study indicate that open-source datasets from Kaggle and Brain MRI Images for Brain Tumor Detection are the most used datasets. However, limited data and imbalanced class problems remain common challenges across various datasets. To overcome those challenges, using a larger dataset, oversampling, Generative Adversarial Network (GAN), federated learning, and Self-Supervised Learning (SSL) to handle the imbalance are the potential solution. Additionally, popular CNN architectures for brain tumor classification extensively use pre-trained models such as VGG16, VGG19, DenseNet121, DenseNet201, GoogleNet, ResNet-50, and Inception-v3. TL strategies are preferred, allowing CNNs to leverage knowledge from large datasets, improving generalization even with limited labeled data.
Blockchain has been relevant in the document management process, serving as a storage solution with the potential to guarantee the relevant requirements needed for any document storage and validation solution. However...
详细信息
ISBN:
(数字)9781665495387
ISBN:
(纸本)9781665495394
Blockchain has been relevant in the document management process, serving as a storage solution with the potential to guarantee the relevant requirements needed for any document storage and validation solution. However, due to the distributed nature of blockchain, we may face implementation difficulties and high operational costs, for example. To facilitate this process, we propose a customizable blockchain-based document registration service that makes it possible to create different types of gen-eralized documents for various application domains and store them in one or more blockchains integrated in an Application programming Interface (API).
Although Twitter is a popular platform for social interaction analysis and text data mining, it faces challenges with geolocation automation. To address this problem, the researchers propose the utilization of a Suppo...
详细信息
Cervical cancer has been known as one of the most prevalent medical disorders globally and a leading cause of death. Early detection, particularly through Pap tests, plays a vital role in its prevention. Previous stud...
Cervical cancer has been known as one of the most prevalent medical disorders globally and a leading cause of death. Early detection, particularly through Pap tests, plays a vital role in its prevention. Previous studies have leveraged machine learning and deep learning techniques to classify the medical images obtained from Pap tests. In this study, a Systematic Literature Review methodology was used to examine 15 relevant papers that have been filtered from queries to Google Scholar which have gone through 4 stages of filtering that include: identification, screening, eligibility, and inclusion. This study addresses two research questions regarding the datasets and deep learning techniques for classifying pap smear images in recent years. The performance of the models was analyzed and potential areas for improvements are suggested. The findings of this study reveal that the Herlev University Hospital and SIPaKMed datasets are the most used. The methodologies used by researchers range from machine learning techniques, transfer learning using Convolutional Neural Networks, and utilize state-of-the-art models with novel optimizing methodology. While there are exciting opportunities in the field, challenges include model generalization and interpretability.
Robots capable of transporting objects are suitable for many applications with societal and economic impact, such as waste retrieval, disposal, and object manipulation in space or the deep sea. However, formulating a ...
详细信息
ISBN:
(数字)9781665462808
ISBN:
(纸本)9781665462815
Robots capable of transporting objects are suitable for many applications with societal and economic impact, such as waste retrieval, disposal, and object manipulation in space or the deep sea. However, formulating a coherent action plan is not trivial due to the size of the search space and the object's physical properties. With the recent advances in Deep Reinforcement Learning (DRL), in this work, we propose, implement, and deploy value-based Deep Reinforcement Methods to tackle the determination of high-level actions that form robust strategies combined with a Probabilistic Roadmap (PRM) method for object transportation through complex environments. The solution was evaluated in a simulation environment and deployed into a real robot. Our results show that DRL can learn strategies effectively, and the robot was able to accomplish its task.
In agricultural water research, the adoption of Internet of Things (IoT) technology has emerged as a pivotal approach for large-scale data collection. Water availability in the context of water quality is very importa...
详细信息
This systematic literature review explores the application of transformer models in early detection of human depression, encompassing text, audio, and video data modalities. Transformer architectures, notably BERT for...
This systematic literature review explores the application of transformer models in early detection of human depression, encompassing text, audio, and video data modalities. Transformer architectures, notably BERT for text, have proven adept at capturing crucial contextual and linguistic patterns associated with depression. For audio and video data, hybrid approaches that combine transformer models with other architectures are prevalent. Key features considered include eye gaze, head pose, facial muscle movements, and audio characteristics such as MFCC and Log-mel Spectrogram, along with text embeddings. Performance comparisons underscore the superiority of text-based data in consistently delivering the most promising results, followed by audio and video modalities when utilizing transformer models. The fusion of multiple modalities emerges as an effective strategy for enhancing predictive accuracy, with the amalgamation of audio, video, and text data yielding the most precise outcomes. However, it is noteworthy that unimodal approaches also exhibit potential, with text data exhibiting superior performance over audio and video data. Nevertheless, several challenges persist in this research domain, including imbalanced datasets, the limited availability of comprehensive and diverse samples, and the inherent complexities in interpreting visual cues. Addressing these challenges remains imperative for the continued advancement of depression detection using transformer-based models across various modalities.
Context: Software development teams constantly opt for faster, lower quality solutions to solve current problems without planning for the future. This situation will have a negative long-term impact and is called tech...
详细信息
Forced phonetic alignment (FPA) is the task of associating a given phonetic unit to a timestamp interval in the speech waveform. Phoneticians are able mark the boundaries with precision, but as the corpus grows it bec...
详细信息
暂无评论